Ever since I embraced low-code development platforms (LCDPs) back in 2016, I've had my fair share of debates about their merits. This includes enterprise architects that are convinced LCDPs could never be enterprise grade, peers that dismiss low-code as a passing fad, and customers wondering if they're just another shiny tool destined to collect dust after the next IT reorg.
Lately, the debate has taken a different turn and we're seeing a whole new wave of articles loudly proclaiming that low-code is dead and being replaced by AI copilots and generative app builders. While these tools are undoubtedly changing the game, my two cents is that low-code isn’t dying; it’s evolving.
Figure 1: Charles Lamanna Addresses the AI Elephant in the Room at the Microsoft Power Platform Community Conference in October
In an enterprise world full of SaaS sprawl, hybrid data landscapes, and disconnected user experiences, low-code platforms have emerged in many organizations as the glue that holds everything together. In that respect, the rise of AI is actually supercharging this trend, making low-code more powerful, more intelligent, and more essential than ever before.
Today’s so-called “vibe coding” and prompt-driven app creation experiences aren’t a wholesale replacement for low-code. Instead, I would argue that they’re the next chapter in their evolution. Whether you’re sketching out a data flow with plain English, generating a workflow from a Teams chat, or building a Copilot that talks to both your CRM and ERP systems, the focus is on rapid application development (RAD). The interface may have changed/evolved, but the goal hasn’t: to make it faster and easier to turn ideas into working software that actually moves the business forward.
In this article, we’ll explore how low-code is evolving alongside AI to power the next wave of enterprise innovation — from integrating hybrid systems and taming SaaS sprawl to modernizing business processes and delivering seamless user experiences — all while maintaining the core foundations and governance needed to keep innovation sustainable.
Where We're Going, We Don't Need Drag-and-Drop
One of the main themes in the history of computer science is the continuous search for a better abstraction. Over the years, breakthroughs have largely come from standing on the shoulders of giants in the field who have developed various types of abstractions that make it way easier to build ever more complex and elaborate solutions.
Although historically most of these innovations have come in the form of richer and more powerful programming languages, more recently the story has been about code generation. In the early days of low-code, vendors introduced graphical tools that made it easier for makers from a wide variety of backgrounds to rapidly develop powerful solutions. Of course, while it may have been simple drag-and-drop on the front end, it was still complex code being generated behind the scenes.
Figure 2: Building an Application Screen Using Microsoft Power Apps
Nowadays with the explosion of generative AI technology, a whole new class of copilot-driven development tools have emerged. Initially, these tools were integrated with pro code developer environments like Visual Studio and, as you can see in Figure 3 below, they provided a huge productivity boost to developers in day-to-day development tasks.
Figure 3: Working with GitHub Copilot
As is the case with pretty much everything in the AI world we live in, the capabilities of these tools have exploded over the past 12-18 months. So much so that we're moved way past "help me write some code to validate a phone number" to "build a time entry app which allows users to record time against a project". The new generative pages in Microsoft Power Apps shown in Figure 4 below are a perfect example of this evolution.
Figure 4: Working with Generative Pages in Microsoft Power Apps
At the end of the day, whether we use a graphical tool or a prompt, the end result remains the same: generated code hosted in a collaborative environment that supports further refinement from both citizen makers and professional coders alike. While it might not be your father's brand of low-code, it's still low-code.
Figure 5: Whether it's Point-and-Click or Prompt Engineering - It's All About Code Generation
Agents Need a Home — LCDPs Provide It
According to a recent study by Zenity, the average large enterprise organization is juggling approximately seven copilots and low-code platforms, spawning nearly 80,000 apps — more than 150 times the number of SaaS apps they're already managing.
While the number of solutions might be much more manageable for smaller organizations, the point is that the coverage area for these types of solutions can (and will) grow in a hurry and you need an environment that can not only host these solutions, but also support you with governance and security.
So, even if you're ready to leave the drag-and-drop low-code behind for the wild frontier of "vibe coding", the reality is that LCDPs represent the stable foundation you need to be successful in getting to where you want to go.
Low-Code Has Never Been a Zero-Sum Game
To see matters clearly, we have to look beyond the marketing hype that has clouded the conversation. Given the natural divide that exists in many organizations between business and IT, it was all too easy for marketing and sales teams to position LCDPs as a silver bullet—a way for business users to bypass IT altogether and build the next generation of enterprise solutions on their own. That messaging may have sold a lot of licenses, but it also created unrealistic expectations on both sides and left a lingering bad taste in everyone’s mouth.
Citizen makers have absolutely proven that they can deliver real value and build meaningful apps faster than ever before, but they still need the support of IT and professional developers to scale, secure, and govern those solutions. When properly embraced, LCDPs provide an ideal environment for citizen makers and professional developers to collaborate on solutions while playing to their respective strengths.
While the ratio of low-code to pro-code can vary from organization to organization, the point is that LCDPs expand the development ecosystem to include a much wider range of people—spanning both functional and technical expertise. Professional developers can use LCDPs to build enterprise-grade solutions dramatically faster, while citizen makers can drive process automation and productivity across departments by building custom apps, workflows, and even AI-powered agents that solve day-to-day problems.
The real power emerges when everyone collaborates within the same environment using fusion teams and generative AI-powered tools are taking that collaboration to the next level. When makers in both camps share components, data models, and best practices, the result isn’t competition, it’s acceleration. That’s the essence of the low-code movement: not replacing IT or reinventing software development, but amplifying the collective creativity of everyone involved. Low-code has never been a zero-sum game. While used correctly, it’s a force multiplier.
Figure 6: Understanding the Benefits of Fusion Development
Closing Thoughts
Low-code development has always been about empowerment — giving people closer to the work the ability to shape the tools they use every day. What’s changing now is the scale and intelligence of that empowerment. AI copilots, natural language interfaces, and agentic frameworks are lowering the barriers to creation even further, making software development more accessible than ever. But with that accessibility comes responsibility. As organizations embrace these new forms of “vibe coding” and AI-assisted development, governance can’t be an afterthought. It must evolve right alongside the tools themselves.
At the end of the day, we shouldn't look at low-code and AI as competing technologies. When mixed properly, they can become complementary forces in a shared mission to make enterprise software more human, more connected, and more responsive. The organizations that will thrive in this new era aren’t the ones that chase every new platform, but the ones that establish a strong foundation for collaboration between IT, the business, and now, intelligent agents. Low-code isn’t dead — it’s getting smarter. And the smartest thing enterprises can do is build the guardrails today that will let innovation run safely and sustainably tomorrow.
